Reckless Endangerment, by Gretchen Morgenson and Joshua Rosner, regurgitates AEI talking points and narratives that were discredited by the Financial Crisis Inquiry Commission.
Imitation is the sincerest form of flattery, so Peter Wallison sang the praises of Reckless Endangerment by Gretchen Morgenson and Joshua Rosner. "I was delighted with the Morgenson/Rosner epic because it was the first book that really supported the position that I have been taking on Fannie and Freddie all along," he told Bloomberg.
Of course Reckless Endangerment supported Wallison's position. The book, which cannot withstand cursory fact checking, repackages the same falsehoods and distortions first promoted by Wallison for The American Enterprise Institute. So far, the authors have escaped critical rebuke by concealing the origins of much of their source material. Perhaps their most transparent deception was Morgenson's claim that, “We have identified as many of our sources as we could when we could."
It seems shocking Morgenson would stain her reputation by writing a book with so many glaring falsehoods. The New York Times reporter won a Pulitzer Prize in 2002 for her "trenchant and incisive" coverage of Wall Street and in The Nation Dean Starkman declared her to be the most important financial journalist of her generation. The book has adulatory blurbs from Bill Moyers and from Charles Ferguson, the director of the Academy Award winning documentary Inside Job. It also received favorable reviews from Robert Reich and from a professor of business journalism. Still, none of these high profile endorsements obviate the litany of journalistic shortcomings in Reckless Endangerment.
Wallison argues that the book refutes the final report by the Financial Crisis Inquiry Commission. FCIC Chair Phil Angelides, “cites the old saw that Fannie and Freddie followed Wall Street into low quality mortgage lending, not the other way around,” he writes. “This is nothing but an urban myth, and is thoroughly disproved not only by Reckless Endangerment, but by the facts in my dissent.” In fact, FCIC eviscerated the central thesis of the Reckless Endangerment, just as it eviscerated the central premise of Wallison’s strident dissent.
The False Equivalency Behind $1.6 trillion of “Toxic Mortgages”
The only "hard evidence" presented by Morgenson and Rosner in support of this thesis--the book relies heavily on innuendo and unsupported assertions--is found in a single sentence:
“By 2008, some $1.6 trillion of toxic mortgages, or almost half that were written, were purchased or guaranteed by Fannie and Freddie.” (p. 117)
This statement is not remotely true. And no one who follows Fannie and Freddie would find that claim to be remotely plausible. Where did that $1.6 trillion number come from? From none other than Peter Wallison, who in November 2008 wrote:
"All told, Fannie and Freddie probably hold or have guaranteed $1.6 trillion in subprime and Alt-A mortgages today."
But Wallison's number was derived from the AEI's unique and fanciful definitions of "subprime" and "Alt-A," which are not used by anyone unaffiliated with a right wing think tank. This sleight of hand--wherein the AEI’s definitions of “subprime and Alt-A” morphed into "toxic”—goes to the deceptive core of the book. It fabricates a false equivalency that goes something like this:
Fannie Mae = low income borrowers = risky subprime mortgage loans = real estate bubble and financial crisis.
This false equivalency it deserves closer scrutiny. It was derived from a series of dubious AEI syllogisms, as follows:
A. Since almost half the mortgages are subprime or some risky equivalent, and
B. Since more than half of all mortgages are guaranteed or owned by Fannie and Freddie, then
C. Fannie and Freddie hold a big chunk of risky mortgages.
That $1.6 trillion number was first calculated by Edward Pinto, a former Fannie executive who became an AEI senior fellow in 2008. In December 2008 he told Congress that everyone else was using the wrong definitions of subprime and Alt-A, and therefore everyone else was undercounting the number of subprime and Alt-A loans in the marketplace and the aggregate subprime and Alt-A risk exposure outstanding.
And, viola! Using Pinto’s definitions, the number of subprime and Alt-A loans more than doubled, from 10.5 million to 25.1 million. (Pinto later increased the subprime/Alt-A number to 27 million, out of a national total of 55 million.) But the risk exposures at Fannie and Freddie exploded. Whereas the GSEs said they owned or guaranteed $8 billion worth of subprime mortgages, Pinto calculated a number that was 80 times higher, about $646 billion. Pinto also said that Fannie and Freddie said they owned or guaranteed another $743 billion in high-risk loans. Lump in the $200 billion in AAA private label mortgage securities held by Fannie and Freddie, and you come up with $1.6 trillion. Pinto argued that Fannie and Freddie financed 34% of all subprime loans and 60% of all Alt-A loans in the marketplace, as of June 30, 2008. Under Pinto’s definitional scheme, the number of non-prime loans held or guaranteed by Fannie and Freddie doubled, from about 5 million to 10 million.
The False Equivalency Behind Subprime Mortgages And Affordable Housing Goals
So how did it come to be that lenders extended high-risk 25.1 million subprime and Alt-A loans? The singular Wallison/Pinto/Morgenson/Rosner explanation is: Affordable housing goals. As Wallison wrote:
The effort to reduce mortgage lending standards was led by the Department of Housing and Urban Development through the 1994 National Homeownership Strategy, published at the request of President Clinton….Once the standards were relaxed for low-income borrowers, it would seem impossible to deny these benefits to the prime market. Indeed, bank regulators, who were in charge of enforcing CRA standards, could hardly disapprove of similar loans made to better-qualified borrowers.
Or as Morgenson/Rosner put it:
How Clinton’s calamitous Homeownership Strategy was born, nurtured, and finally came to blow up the American economy is a story of greed and good intentions, corporate corruption and government support. (p. 6)
Which brings us to the second AEI syllogism:
A. Since Fannie and Freddie expanded their affordable housing goals,
B. They expanded the market for subprime loans, and consequently
C. The GSEs led the race to the bottom in credit standards for mortgage loans, ergo
D. They created the financial crisis of 2008.
Wallison, Pinto, Morgenson, and Rosner all claim that Fannie led the way. “In 1994, Fannie Mae replaced its initial $10 billion program with a $1 trillion affordable housing initiative,” writes Wallison in, “The Last Trillion-Dollar Commitment.” Morgenson and Rosner ascribe the lion’s share of the blame to James Johnson, Fannie’s CEO from 1991 to 1998. They write:
In March 1994, Johnson announced Fannie Mae’s Trillion Dollar Commitment, a program that earmarked $1 trillion to be spent on affordable housing between 1994 and 2000…“This was the beginning of the world ‘trillion’ in housing,” marveled Edward Pinto, a former Fannie executive. “When Johnson announced the $1 trillion commitment, it told Washington that the company could do more than they were already doing. It was pouring gasoline on the fire.” (p. 59)
As they had explain in the opening of the book:
Johnson's tactics were watched closely and subsequently imitated by others in the private sector, interested in creating their own power and profit machines. Fannie Mae led the way and relaxing loan underwriting standards, for example, a shift that was quickly followed by private lenders. Johnson's company also automated the lending process so that loan decisions could be made in minutes and were based heavily on the borrower's credit history, rather than on a more comprehensive financial profile that had been the case in the past.
Eliminating the traditional due diligence conducted by lenders soon became the playbook for financial executives across the country. Wall Street, always ready to play the role of enabler, provided the money for these dubious loans, profiting mightily. Without Wall Street firms giving billions of dollars to reckless lenders, hundreds of billions of bad loans would never have been made. (p. 5)
[…]
“Clinton was clearly coordinating with [Johnson]—they had the same goals at the same time,’ said Edward Pinto, former chief credit officer at Fannie Mae, who is a consultant. (p.11)
Morgenson and Rosner never disclosed Pinto’s affiliation with the AEI. Nor did they mention that he was fired from his credit position in 1989, after less than two years on the job, before Johnson joined the firm.
Here’s how Morgenson and Rosner explain how Fannie and Freddie got saddled with $1.6 trillion in “toxic mortgages”:
Regulators were relaxing rules, Fannie and Freddie were inflating their portfolios, homebuilders and subprime lenders were flexing their muscles too. To meet the goals Fannie and Freddie had to buy riskier mortgages, such as those defined as subprime.
Some $160 billion in subprime loans would be underwritten in 1999, up from $40 billion five years earlier. And in another four years, that figure would jump to $332 billion.
Many of those loans wound up in Fannie's and Freddie's portfolios. By 2008, some $1.6 trillion of toxic mortgages, or almost half that were written, were purchased or guaranteed by Fannie and Freddie.
In 1999, the growth trajectory that [former Fannie CEO James] Johnson had dreamed of was fast becoming a reality.
Except that growth trajectory was not dreamt up by James Johnson, it was fabricated by Morgenson and Rosner, who, once again, rely on that old trick of false equivalency. Notice how subprime loans grow from $160 billion in 1999, to $332 billion in 2003, and then in 2008 Fannie and Freddie held $1.6 trillion in “toxic loans.” Morgenson and Rosner don’t reveal where their numbers come from, but they show up in one of Pinto‘s reports (p. 124). The subprime totals were taken from a Standard & Poor’s publication, Inside B&C Lending. (At one time, “B&C” loans were considered synonymous with subprime loans, though precise definitions were in the eye of the beholder.)
The conflation between these B&C loans and the mortgages extended by the GSEs would not pass the laugh test. Nor could anyone honestly argue that the regulators forced Fannie and Freddie to buy risky B&C loans to meet government housing goals. Specifically:
[W]ith respect to the subprime market, HUD believes that the risky, B&C portion of that market should be excluded from the market estimates for each of the housing goals. Thus, HUD includes only the A-minus portion of the subprime market in its overall estimates of the goals-qualifying market shares.
Morgenson and Rosner simply follow the AEI party line, which states that federal regulators promoted a reduction in credit standards, and they ignore all evidence to the contrary, such as HUD’s anti-predatory lending guidelines, which precluded the GSEs from taking the types of loans that were pervasive among private subprime lenders. HUD expressly prohibited the inclusion of “Mortgages contrary to good lending practices,” such as those with:
Evidence that the lender did not adequately consider the borrower’s ability to make payments, i.e., mortgages that are originated with underwriting techniques that focus on the borrower’s equity in the home, and do not give full consideration of the borrower’s income and other obligations. Ability to repay must be determined and must be based upon relating the borrower’s income, assets, and liabilities to the mortgage payments.
HUD regulations constrained the GSEs from extending loans high upfront fees and prepayment penalties, which were standard in the private label subprime market. As we’ll see, the differences between the terms of GSE and private label subprime mortgages corresponded to the differences in loan performance.
The Disconnect Between “Toxic Mortgages” And Loan Performance
Johnson's tenure at Fannie began in 1990 and ended 1998, yet Morgenson and Rosner insist that Johnson’s actions caused severe damage that only became apparent one decade after he left his job. So how did Johnson and Fannie do? The answer is summarized each year by the Federal Housing Finance Agency in its Annual Report to Congress.
Check out “Table 8. Fannie Mae Mortgage Asset Quality,” in the 2010 Report to Congress, which shows how Fannie’s loans have performed over the past 35 years. For two decades, from 1988 through 2007, Fannie's serious delinquency rate for single family mortgages was consistently below 1%, and actual credit losses were a small fraction of those delinquencies. In 2008 and thereafter, serious delinquency rates were far worse, peaking at 5.38% in 2009. And credit losses as a percentage of the total book reached an all time high in 2010, when they averaged 0.77%. That's right, after the worst housing crash in U.S. history, the loss rate on Fannie's overall mortgage book hovers around 1%.
It’s simply impossible to argue, as Wallison, Pinto, Morgenson, and Rosner do, that Fannie and Freddie led the race to the bottom in credit standards, because in terms of actual loan performance over the past decade, the GSEs are vastly superior to any other segment of the residential mortgage market. Even Bush appointee James Lockhart, a tireless critic of the GSEs, confirmed this truism in an AEI presentation.
Which brings us back to the scurrilous claim that: “By 2008, some $1.6 trillion of toxic mortgages, or almost half that were written, were purchased or guaranteed by Fannie and Freddie.” At no time, during 2008 or any time thereafter, did Fannie and Freddie ever hold or guarantee $1.6 trillion in troubled loans. And $1.6 trillion was not to “close to half,” of Fannie and Freddie's combined loan total of $4.5 trillion; it was one more like one-third. Fannie and Freddie’s delinquency rates peaked in January 2010, when all delinquencies, even those 30 days late, represented 8.5% of the total loans outstanding.
Here’s what the FCIC had to say:
The Commission also probed the performance of the loans purchased or guaranteed by Fannie and Freddie. While they generated substantial losses, delinquency rates for GSE [i.e. government sponsored enterprise] loans were substantially lower than loans securitized by other financial firms. For example, data compiled by the Commission for a subset of borrowers with similar credit scores—scores below 660—show that by the end of 2008, GSE mortgages were far less likely to be seriously delinquent than were non-GSE securitized mortgages: 6.2% versus 28.3%. [The FHFA reported that a wide disparity continued through 2010.]
We also studied at length how the Department of Housing and Urban Development’s (HUD’s) affordable housing goals for the GSEs affected their investment in risky mortgages. Based on the evidence and interviews with dozens of individuals involved in this subject area, we determined these goals only contributed marginally to Fannie’s and Freddie’s participation in those mortgages.
The FCIC also examined Pinto’s research at length, and conclusively demonstrated back in August 2010 that his categorization scheme made no sense. As the final report states:
Importantly, as the FCIC review shows, the GSE loans classified as subprime or Alt-A in Pinto’s analysis did not perform nearly as poorly as loans in non-agency subprime or Alt-A securities. These differences suggest that grouping all of these loans together is misleading. In direct contrast to Pinto’s claim, GSE mortgages with some riskier characteristics such as high loan-to-value ratios are not at all equivalent to those mortgages in securitizations labeled subprime and Alt-A by issuers. The performance data assembled and analyzed by the FCIC show that non-GSE securitized loans experienced much higher rates of delinquency than did the GSE loans with similar characteristics.
The FCIC review was echoed by a comprehensive analysis performed by the FHFA. What do Wallison/Pinto/Morgenson/Rosner say about any of this? Absolutely nothing. Or rather, Wallison‘s FCIC dissent argued, rather incredibly, that the FCIC never reviewed Pinto’s work.
The Disconnect Between Johnson’s Track Record and The Track Records Of Others
What about Fannie’s post-2007 losses, which pushed it into insolvency? Can those be plausibly traced to James Johnson, as Morgenson and Rosner contend? Wallison argues in The Wall Street Journal that the FCIC overlooked James Johnson’s role, and therefore the investigation was a whitewash. “Mr. Johnson was not among the more than 700 witnesses the commission claims to have interviewed,” he writes.
In fact, the FCIC published a preliminary report in April 2010 demonstrating how Fannie’s standards began their downward slide back in 2004, six years after Johnson left the company. For instance, by year-end 2004, Fannie held $36 billion in interest only mortgages. Three years later, that number increased to $207 billion. Though this loan product represented only 8% of the single family mortgage book, it represented 33% of all credit losses in 2008 and 2009. By definition, interest only mortgages would be considered to be “contrary to good lending practices,” and therefore ineligible for affordable housing goals. Interest only and Alt-A loans represented, together, about 16% of Fannie and Freddie’s mortgage portfolios, but they contributed the majority of the credit losses. Once again, the Morgenson/Wallison claim that the “toxic mortgages” were “almost half that were written,” turns out to be bunk.
Morgenson and Rosner don’t let facts interfere with their preexisting narrative about Johnson and Fannie Mae. Recall this passage:
Johnson's tactics were watched closely and subsequently imitated by others in the private sector, interested in creating their own power and profit machines. Fannie Mae led the way and relaxing loan underwriting standards, for example, a shift that was quickly followed by private lenders. Johnson's company also automated the lending process so that loan decisions could be made in minutes and were based heavily on the borrower's credit history, rather than on a more comprehensive financial profile that had been the case in the past.
It’s true that private lenders followed the lead of the GSEs in using financial models for automated underwriting. But it was Freddie, not Fannie, which led the way in promoting the use of automated underwriting programs for nonconforming loans. And Freddie’s effort was a pilot program, a very small part of the GSEs’ overall business or the overall market. Representing a small pilot program initiated by the GSEs as a sea change in the overall mortgage markets is yet another device for creating a false equivalency. On May 10, 1996, American Banker reported:
Freddie Mac says its first significant foray into the non-conforming loan market will revolutionize the industry for these credits. The housing agency predicts more banks will offer the loans, underwriting will change, and consumers will pay lower rates.
"This represents a major step forward," said Peter Maselli, vice president of automated underwriting at Freddie Mac, formally the Federal Home Loan Mortgage Corp. Mr. Maselli was discussing a new program in which an affiliate of Countrywide Credit Industrie will buy nonconforming loans that lenders have evaluated using Freddie Mac's software.
…
For its part, Fannie Mae, which also has an automated underwriting system, plans to steer clear of involvement with B and C loans, a spokesman said. He added, however, that the agency is working on a program for jumbo loans that would operate along the lines of the Freddie Mac initiative. The spokesman said Fannie expects to roll out the program later this year.
Parallels in Reporting That Strain Credulity
It’s hard to believe that Morgenson and Rosner did reporting that was independent of the AEI when you compare their analysis of an Urban Institute report with that of Peter Wallison in his article, “Cause and Effect.” They all misrepresent the report as a turning point in the degradation of GSE lending standards, and they all misrepresent the report in precisely the same way. So while these falsehoods do only minor damage to the book’s overall narrative, they raise damning questions about the authors’ credibility.
First consider Wallison’s text, which was written in November 2008:
In 1997, for example, HUD commissioned the Urban Institute to study the GSEs' underwriting guidelines. The Urban Institute's report concluded: "The GSEs' guidelines, designed to identify creditworthy applicants, are more likely to disqualify borrowers with low incomes, limited wealth, and poor credit histories; applicants with these characteristics are disproportionately minorities. Informants said that some local and regional lenders serve a greater number of creditworthy low-to-moderate income and minority borrowers than the GSEs, using loan products with more flexible underwriting guidelines than those allowed by Fannie and Freddie."
Following this report, Fannie and Freddie modified their automated underwriting systems to accept loans with characteristics that they had previously rejected. This opened the way for the acquisition of large numbers of nontraditional and subprime mortgages. These did not necessarily come from traditional banks, lending under the CRA, but from lenders like Countrywide Financial, the nation's largest subprime and nontraditional mortgage lender and a firm that would become infamous for consistently pushing the envelope on acceptable underwriting standards.
The report did not “conclude” what Wallison said it concluded. Nor did it recommend any relaxation in credit standards. The report’s conclusion was that the matter warranted further study, which could be obtain through better information gathering. Wallison presents no evidence that the GSEs relaxed their credit standards in response to the report, and his conflation Countrywide’s subprime loans, which were sold through private label securitizations, and its sale of conforming loans to the GSEs, is yet another eample of the false equivalency that runs through all his writings.
Now consider the text from Reckless Endangerment:
HUD [requested] that the Urban Institute, a nonpartisan and nonprofit economic policy and research organization, conduct research on the current efforts by Fannie and Freddie to finance mortgages for low-income people in “underserved” communities. The report, “A Study of the GSEs’ Single-Family Underwriting Guidelines,” was funded by HUD and published in April 1999.
The report had four authors: Kenneth Temkin, George Galster, Roberto Quercia, and Sheila O'Leary. They canvassed “knowledgeable observers” in the mortgage and housing field in Boston, Detroit, Miami, and Seattle; all had experience dealing with Fannie and Freddie.
While acknowledging that Fannie and Freddie did a fine job financing many home mortgages and noting that the companies’ new, more flexible loans helped borrowers who had been previously shut out of the housing market, the authors nevertheless concluded that Fannie and Freddie were not doing enough for low-income borrowers.
"The GSEs' guidelines, designed to identify creditworthy applicants, are more likely to disqualify borrowers with low incomes, limited wealth, and poor credit histories; applicants with these characteristics are disproportionately minorities,” the study said. “Informants said that some local and regional lenders serve a greater number of creditworthy low- to moderate-income and minority borrowers than the GSEs, using loan products with more flexible underwriting guidelines than those allowed by Fannie Mae or Freddie Mac.”
The message was clear: Fannie and Freddie’s underwriting standards were too high; low-income originations were too low. (p. 115)
The “message” of the report was invented by Morgenson and Rosner. When you read the “Results,” section of the report in context, it’s obvious that the informed sources had stated that the GSEs’ credit standards were higher than some private lenders, and that higher credit standards tend to disqualify low income and minority borrowers. But here’s the section of the report that discredits the interpretations made by Wallison, Morgenson and Rosner. It states that these informed sources offered no evidence to challenge the validity of the GSEs' higher standards:
These reports may mean that some local and regional primary lenders are using guidelines that are less likely to disqualify minority mortgage applicants and that accurately predict loan performance. This raises the possibility that the GSEs' current guidelines may have a disparate impact on minority borrowers if the guidelines have disproportionate effects on minority borrowers but do not serve a business necessity by accurately predicting loan performance. However, none of the informants provided the evidence necessary to confirm this possibility. A disparate impact finding would have to be based on a detailed analysis of the relationship between underwriting guidelines and loan performance, but none of the informants presented any specific empirical data about the performance of loans issued with underwriting guidelines more flexible than the GSEs'.
Consequently, the real message of the report said nothing about lowering underwriting standards:
Recommendations
In order for HUD to systematically address the concerns raised by the informants, we recommend that HUD (1) develop an enhanced loan-level database that includes loan performance information for mortgages purchased by the GSEs and that is supplemented, to the extent possible, with loan-level performance information on other mortgages within the conforming limit; (2) analyze automated underwriting outcomes with hypothetical mortgage applications that contain information on different types of borrowers; (3) improve current simulation models that estimate the effects of possible GSE guideline changes on various types of borrowers; and (4) improve direct communication mechanisms so that lenders and advocates can raise issues with HUD on an ongoing basis.
These steps would improve HUD's ability to determine whether those GSE guidelines disproportionately affecting minority borrowers serve a business necessity. In addition, the recommended steps would allow HUD to assess the impact automated underwriting systems and underwriters' use of credit scores would have on different types of borrowers, as well as helping the department remain up-to-date on issues of concern to lenders.
There are plenty of other distortions in the book, which we may detail in future diaries. After a while, you get a sense of déjà vu, which is how the right wing echo chamber operates. It’s all reminiscent of that other false equivalency from nine years ago:
9/11 = Al Qaeda = terrorism = Saddam Hussein = direct threat to the United States
Dr. Allan Mendelowitz, the former Chairman of the Federal Home Loan Board, put it this way back in November 2008:
There were two big lies in the past eight years. One is that there are weapons of massive destruction in Iraq and Saddam Hussein was associated with Al-Qaeda. That was one. That one got a lot of coverage. The other big lie, that hasn't received quite as much coverage is that somehow Fannie Mae and Freddie Mac were the cause of the crisis that we faced.
And the usual suspects parrot the same party line. "[T]he Fannie Mae scandal is the most important political scandal since Watergate," writes David Brooks. "Fannie was a cancer that helped spread risky behavior and low standards across the housing industry. We all know what happened next."
The authors of the book will be available for questions at an FDL Salon this afternoon.